A sports-prediction network whose product earns a fee on winning trades and channels that revenue into buybacks of its token.
Sportstensor turns a network of competing predictive models into a single, tradable edge on sports outcomes. Miners submit predictions across major leagues such as the NBA, NFL, MLB, and the English Premier League; these are aggregated into a meta-model whose signal is strong enough to trade. That capability is packaged into a product called Almanac, which places positions on prediction markets via Polymarket.
The result is a subnet with a genuine, external source of revenue: it does not depend solely on token emissions to function, because the underlying predictions have value that a market will pay for.
Almanac charges a fee of roughly one percent on winning trades. That revenue is used to buy the Sportstensor alpha token on the open market, with the option to burn a portion of what is bought. Commentators have described it as one of the first direct links between a Bittensor subnet's real-world revenue and its token, which is precisely why it clears our bar.
The design is also deliberately anti-dilutive: the project frames it so that emissions cannot outrun earnings, meaning the token's incentive budget is anchored to what the product actually generates rather than to an open-ended emissions schedule. The buyback became operational around the launch of Almanac's open beta in late 2025.
Sportstensor takes an unusually clean approach to the question we always ask: does revenue reach the token? A fee on winning trades flows into open-market buybacks, and emissions are tied to earnings. It is a small, focused example of the revenue-to-token link we want to own across the network.
A few nuances matter. The fee is levied on winning trades rather than on raw trading volume, so revenue scales with the model's accuracy and the size of the book, not simply with activity. The burn is conditional rather than automatic; the routine, confirmed action is the buyback. Predictive trading is inherently variable, prediction-market regulation is unsettled, and the token is young and illiquid. As with every subnet position, we treat this as high-risk, venture-style exposure and size it accordingly.
Supporting sources: the Sportstensor documentation and a published "Subnet Session" interview with the team; analyst coverage by subnetalpha.ai and IQ.wiki. Live network data: taostats.io/subnets/41. Mechanisms in the dTAO era change quickly; this reflects our reading at the time of writing.